What’s wrong with single-member electorates - Part 2

To limit the advantages accruing to the largest parties because of vote wastage in party-list systems of proportional representation based on average votes per seat won, various Scandinavian countries have continued to refine a number of different elaborate allocation mechanisms. The modified d'Hondt scheme the Federal Parliament inflicted on the Australian Capital Territory in 1989 and 1992 highlighted the arbitrariness and instability associated with qualification thresholds: on both occasions, several of the seventeen seats depended on whether one party finished a few dozen votes above or below the 5.56 per cent threshold!

Other unsatisfactory multi-member-electorate experience has been witnessed in Japan and Australia.

For a long time in Japan, voters simply wrote down the name of one candidate in their multi-member constituency and the candidates with the highest number of votes were elected. Where one candidate was extremely popular, the way was paved for others, not necessarily from the same party, to be elected with rather few votes. More alarmingly, the inability to transfer excess votes to other candidates in the same party fuelled the 'money politics' striving of individual candidates and their backers that spawned a series of major scandals related to political donations.

The Australian Senate became a national laughing-stock because of its regularly lop-sided composition prior to 1949. Initially voters were required to mark as many crosses as there were vacancies and the candidates with greatest support, usually from the same party, were elected. The situation changed little in practice when preferential voting was introduced and vacancies were determined sequentially by ignoring the names of candidates already elected and seeing who next achieved a majority of votes.

These examples illustrate the importance, where fairness of outcome is seen as important, of being able to transfer votes from those who have more than strictly required for election and from those with no longer any hope of election. This is best done by giving voters a single transferable vote in multi-member electorates (their marking of preferences indicates the order in which candidates can have access to that vote) and using a quota-preferential count in which candidates are not required to obtain more votes than necessary to be certain of election.

The (Droop) quota is found by dividing the number of formal votes by one more than the number of vacancies and moving up to the next highest whole number. Surpluses above the quota are available for transfer to others in accordance with individual voters’ wishes. Alternatively, the candidate with fewest votes is excluded and each of his or her ballot-papers is examined to see who of the continuing candidates has highest preference.

The introduction of such a system of proportional representation in 1949 has been pivotal in transforming the Senate into a recognised force for government and public service accountability. However the Senate voting system has two major defects as well as significant technical flaws related to the transfer of surpluses from elected candidates: its way of filling casual vacancies adopted after the 1975 constitutional crisis could also have been much improved by re-examining voters' expressed wishes.

With the enlargement of the national Parliament in 1984, the number of Senate vacancies that usually occur increased from five to six. When six Senators are being elected, just over one-seventh of the vote guarantees anyone election: 42.9 per cent is enough for three places while over 57 per cent is required to guarantee four, and therefore out of reach except in extreme circumstances.

Having an odd number of vacancies is normally important in any quota-preferential election as only then does a majority of votes in an electoral district always translate into a majority of seats.

The Senate voting system also offers safe seats to those at the top of major party columns. Preselections are therefore often fiercely contested in relation to that internal order, and in practice usually only two of the six positions (sometimes three in Queensland because the Liberals and Nationals run separately) are in doubt once nominations have closed.